This application proposes a network to stimulate and foster collaborations that have the potential to lead to new methodology for the development and evaluation of adaptive treatment strategies for chronic disorders. These strategies are individually tailored approaches to treatment that correspond to the adaptive nature of clinical practice. For example, in practice, patients suffering from chronic disorders such as substance abuse, mental illness and HIV infection are treated sequentially, where decisions to modify or change treatment over time are made based on the clinician's and patient's evaluation of ongoing response, burden and adherence. Both data analysis methods and experimental data collection methods for informing and evaluating adaptive treatment strategies are in their infancy. This issue can be viewed from the perspective of computer scientists and control engineers as a multi-stage decision problem; however there has been little interchange between clinicians involved in the management of these disorders, statisticians who work in data analysis and design and computer scientists. Input from all of these areas has the potential to jump-start the methodological development in a manner that will ensure that major challenges will be identified and addressed in the most appropriate fashion. This network includes computer scientists, psychiatrists, psychologists, engineers and statisticians. We will write a joint position paper identifying the major challenges and potential solutions and we will form small collaborative groups that will work to address these methodological challenges.